Nowadays, especially after the great developments in computer systems, prediction and estimation applications using artificial neural networks (ANN) have reached an extremely important point in terms of different sciences and sectors. In the 21st century, prediction ability of image qualifications has led to making a profit in many sectors both in terms of time and number of staff. Prediction can be thought of as the mental process that allows us to predict or develop a response when there is no other solution. Namely, prediction is a nonlinear situation. ANN are frequently used in solving prediction problems, as ANN have proved their success in the whole world, especially in solving nonlinear problems. In this study, a system is modeled by using image processing techniques and ANN to predict the total coin amount in the image.

:  Fuel Cells will play a significant role among the energy storage devices, and it is one of the vital technical challenges in the present scenario. Various technologies such as Direct Methanol Fuel Cell (DMFC), Hydrogen Fuelled Fuel Cell, Borohydride Fuel Cell, PEM Fuel Cell, Nano-material fuel cell are in vogue for optimizing the energy storage. However, cost-effective PEM-based red-ox flow fuel cell systems are highlighted in this work. In the hybrid system, while integrating the various energy sources, it is challenging to integrate them. Thus, a proper design is required to analyze the techno-economic aspects of integrated renewable energy systems. Hence, this work paves the way to optimize the intermittent nature of various renewable energy sources using simulation software tools such as INSEL, TRNSYS, RETScreen, ARES and HOMER are used for optimizing the maximum efficiency of the fuel cell for photovoltaic applications. This paper also addresses the Trade-Off inference of various software tools available for analyzing the integration of renewable energy sources for sustainable energy development. This work has attempted to optimize the Hybrid Renewable Energy Sources using HOMER software for climatic forecasting. The simulation results of the proposed hybrid system are presented using MATLAB/Simulink software.

The demand on telecom companies is still exponentially increasing on daily bases. In parallel, customer complaints from the services will have a similar curve. Most telecom companies rely on customer feedback to evaluate their network and services. In this thesis, we will take a Lebanese telecom company as a case study, and we will study the implementation of Machine Learning algorithms on it, using Artificial Neural Networks (ANN), comparing the feasibility of multiple optimizers and activation function to find the one that best suits our case. We are using a sample database of 10,000 mobile market subscribers with variables of gender, age, device manufacturer, service quality, and complaint status. We also propose the segmented-prediction model by window (interval time) and customer groups for better accuracy and practical usage. The customer group will have examined by gender, age, device manufacturer, and region area.

Visual-based target detection, that is, using some target detection algorithm to identify potential targets in images or videos, is an important research topic in the field of visual target detection. Intelligent vehicles will face various complex scenes in the real traffic environment in the process of driving, which puts forward higher requirements for target detection methods. In recent years, great progress has been made in the field of deep learning. In particular, deep convolutional neural network shows far more advantages than traditional algorithms in the field of computer vision, such as image recognition, target detection and semantic segmentation. Traffic target detection of vehicles, signs and other traffic targets based on deep convolutional neural network has gradually become a new research trend in autonomous driving technology, intelligent transportation system and other fields. Based on the detection principle of Faster R-CNN, this paper conducts training on the samples of signs and vehicles. The results show that the traffic target detection based on Faster R-CNN is effective, with high detection accuracy and fast speed.

Outbreak of Covid-19 and related its progression across the extensive landscape around the globe affecting a humongous extent of population at various geographical locales have posed a state of pandemic, hard to quantify of its extensiveness, temporal progression, and spatial proliferation.  Psychological stress and panic syndrome caused by the virility of Corona virus have posed a query on the uncertainty aspects of its spatiotemporal spreading in human society.  Assessing the associated etiology, diagnostic schemes, therapeutic regimens, and prevention methods etc. is explicitly vital and obviously needs quantifiable trends or forecasting techniques. It goes without saying that such schemes have great importance in present situation and as such, conceived and elaborated in this study is a strategy towards forecasting the temporal trend of the disease-progression on ex post basis using the observed ex ante details word wide from the onset of pandemic outbreak of Covid-19. The associated stochastic modelling of the proposed forecast method is based on artificial neural network (ANN) with feasible extensions of deep-learning heuristics. This paper outlines the challenges involved thereof in performing forecasting of widely spread outbreaks specified in vagaries of timeslots and across ensembles of incidence plus mortalities. The underlying state of evolution is modelled using ergodic profiles of pandemic state of the disease in question; in all, the scope of this study is to forecast temporal progression of Covid-19 pandemic in two chosen countries via a fast-convergent ANN yielding ex post predictions (on number of mortality profiles) consistent with ex ante data availed.